Convolutional-Neural-Networks-Theory

Convolutional networks(LeCun, 1989), also known as convolutional neuralnetworks, or CNNs, are a specialized kind of neural network for processing datathat has a known grid-like topology. Examples include time-series data, which canbe thought of as a 1-D grid taking samples at regular time intervals, and image data,which can be thought of as a 2-D grid of pixels. Convolutional networks have beentremendously successful in practical applications. The name “convolutional neuralnetwork” indicates that the network employs a mathematical operation calledconvolution. Convolution is a specialized kind of linear operation. Convolutionalnetworks are simply neural networks that use convolution in place of general matrixmultiplication in at least one of their layers. We describe several variants on the convolution function thatare widely used in practice for neural networks. We also show how convolutionmay be applied to many kinds of data, with different numbers of dimensions. We then discuss means of making convolution more efficient. If you want to know more go through this link - https://www.kaggle.com/discussions/general/395941